Grading changes one year on

[Alex Holowczak]

For the Warwickshire Junior team for this weekend, I've been working out their grades this season compared to last season. I can only use League chess for this, since they're the only things the results are kept for, and I have to hand. So there are no Congresses, 4NCL or anything. So I have the results of the Birmingham, Wolverhampton, Cannock and Leamington Leagues this season, worked out their grades, and compared them to the new grades they were issued last season. Nearly all players play in these, so it seems useful to use to compare to one another.

This information is presented in the format: Name Grade-League Grade This Year (Change)

Richard Weaving 181-187 (+6)
Kaiser Malik 180-179 (-1)
Henrik Stepanyan 169-159 (-10)
Tom Lilburn 165-151 (-14)
Robert Statham 160-142 (-18)
John Gilbert 150-130 (-20)
Manpreet Sangha 143-158 (+15)
Hari Dhesi 143-138 (-5)
Philip Purcell U-142 (NA)
Shakeel Choudhury 135-156 (+21)
Philomena Lip 124-NA (NA)
Arron Dhesi 110-96 (-14)

That's a net loss of 40 points for the 10 players a difference can be calculated for, i.e. 4 points per player. There's a healthy mix of ages here too, between 13 and 18, I can't draw any connection between age and grade alteration. The only thing I can think of is where the players play their chess. The top 6 there all played in Division One of their leagues, whereas the others played in lower divisions, where the standard of opposition was naturally lower. In which case, they'll jump up next year into higher divisions of their leagues, so they might come crashing down again.

Usually, juniors' grades go up every season, but this season, that looks like it won't be the case. Questions arise:
1) Has this been the case in the rest of the country?
2) Is the stretching that had previously happened happening again?
3) Are my means of sampling (by ignoring 4NCL, County and Congresses) sensible?

For the record, my grade looks set to go from 158-139 (results against ungradeds may mess that up), which means my grade is now below Sangha and Purcell, who were always below me in the school team I used to play for.

[Roger de Coverly]

Those of us who are direct members may have received our provisional grading printout. Leagues may not be included, but I can see the pattern of junior grades reducing from one year to the next being repeated in the small sample of those I played last season.

The graders had this belief that the gap between the "average" player and "GM Keith" and "GM Mark" was too wide by about 20-25 ECF points and therefore did some cooking to try and reduce it. This is notwithstanding the fact that both the "average" player and "GM Keith" seemed to have fairly stable grades from one year to the next. Much of the cooking seemed to involve ignoring the previous grades of junior players. Grading systems by their nature should be self correcting so we may be seeing the "natural" position being re-established.

The junior revaluation issue is interesting. If you have a player assigned a 150 grade who only plays to a 140 standard, then you inflate his or her opponents slightly. If you treat them as a new player, then the inflation effect is removed.

[Angus French]

Alex, can you say how many games each of your juniors played?

FWIW, I played six juniors last year. My preliminary results check list shows them with what I assume are estimated grades. I further assume that these are based on their results, so far received, against graded players for the season just ended. The estimates, if that's what they are, are down 25 points in total on the grades for 2009/10 - an average reduction of just over 4 points per player (the same as the figure derived by Alex).

[Alex Holowczak]

Alex, can you say how many games each of your juniors played?

I found between 9 and 20, but most of those will have played more stuff to take the total over 20. So I don't see why they'd need last year's results to make up the 20.

[Rob Thompson]

Alex, can you say how many games each of your juniors played?

FWIW, I played six juniors last year. My preliminary results check list shows them with what I assume are estimated grades. I further assume that these are based on their results, so far received, against graded players for the season just ended. The estimates, if that's what they are, are down 25 points in total on the grades for 2009/10 - an average reduction of just over 4 points per player (the same as the figure derived by Alex).

Out of interest, can you tell me what my estimated grade is?

I've played 22 juniors, who have gone up by 59 points in total, giving me a boost of about half a point on my performance. However, these are likely to be more active than most, given who i've played. Of course, some have dropped points and some have gained them, but overall my junior opponents have bucked the trend noticed by others

[Alex Holowczak]

Rob, could this be because you, and the people you play against, play so many games against other juniors. I don't know your total number of games, but I'd assume it's about 20% of your total. Conversely, in the case above, all of the games were against adults, not juniors (because I don't have the junior information available to me). So perhaps that has something to do with it too?

[Rob Thompson]

maybe, but unlikely, as most strong juniors play the majority of their games against adults anyway. It's possible though, that it had a small effect. It was 22/103, so nearer 20% of my games

[Angus French]

Alex, can you say how many games each of your juniors played?

FWIW, I played six juniors last year. My preliminary results check list shows them with what I assume are estimated grades. I further assume that these are based on their results, so far received, against graded players for the season just ended. The estimates, if that's what they are, are down 25 points in total on the grades for 2009/10 - an average reduction of just over 4 points per player (the same as the figure derived by Alex).

Out of interest, can you tell me what my estimated grade is?

In my list, your grade - or what I assume is your grade - is shown as *168.

[Alex Holowczak]

In my list, your grade - or what I assume is your grade - is shown as *168.

Which is actually -2 from your grade this season. So this would correspond with my theory that juniors grades have decreased slightly this season.

[Rob Thompson]

that's exactly what I had worked out my grade to be, without the junior uplift added on. Presuming they are still doing that, i should be 173 next year, and many of the other slight drops will be corrected

[Angus French]

In my list, your grade - or what I assume is your grade - is shown as *168.

Which is actually -2 from your grade this season. So this would correspond with my theory that juniors grades have decreased slightly this season.

Quite so.

I had erroneously recorded Rob's grade as 180.

So: my 6 junior opponents are apparently scheduled to lose 15 points (rather than 25).

[Neill Cooper]

I thought that the new junior grade calculation is broken. I think this is shown by the January rapid play grades.
I've played lots of juniors this year and their * grades are mainly a few point below last year's published grades.
Perhaps that shows that the 'predicted' grades last year were slight over predictions of their performance this year.

[Ben Purton]

where can i get my grade from?

[Gareth Harley-Yeo]

where can i get my grade from?

Get mine while you're at it. I could do with a giggle.

[Alex Holowczak]

where can i get my grade from?

If you're a direct member, you ought to have been e-mailed a provisional list of results by now. You'll see all your results (perhaps missing one or two due to slow graders!), and just do the maths.